Spinophilin Stabilizes Cell Surface Expression of α2B-Adrenergic Receptors

The third intracellular (3i) loops of the α2A- and α2B-adrenergic receptor (AR) subtypes are critical for retention of these receptors at the basolateral surface of polarized Madin-Darby canine kidney (MDCKII) cells at steady state. The third intracellular loops of the α2A, α2B, and α2C-AR subtypes interact with spinophilin, a multidomain protein that, like the three α2-AR subtypes, is enriched at the basolateral surface of MDCKII cells. The present studies provide evidence that α2-AR interaction with spinophilin contributes to cell surface stabilization of the receptor. We exploited the unique targeting profile of the α2B-AR subtype in MDCKII cells: random delivery to apical and basolateral surfaces with rapid (t½ ≤ 60 min) apical versus slower (t½ = 10–12 h) basolateral turnover. Apical delivery of a spinophilin subdomain containing the α2-AR-interacting region (Sp151–483) by fusion with apically targeted p75NTR extended the half-life of α2B-AR at the apical surface to ∼3.6 h and eliminated the rapid phase (0–60 min) of α2B-AR turnover on that surface. Furthermore, we examined α2B-AR turnover at the surface of mouse embryo fibroblasts derived from wild type (Sp+/+) or spinophilin knock-out (Sp–/–) mice. Two independent experimental approaches demonstrated that agonist-evoked internalization of HA-α2B-AR was accelerated in mouse embryo fibroblasts derived from Sp–/– mice. These findings are consistent with the interpretation that endogenous spinophilin contributes to the stabilization of α2B-AR and presumably all three α2-AR subtypes at the surface of target cells and may act as a scaffold that could link α2-ARs to proteins interacting with spinophilin via other domains.

␣ 2 -Adrenergic receptors (ARs) 1 are members of the large superfamily of G protein-coupled receptors that contain seven putative transmembrane spanning regions. There are three ␣ 2 -AR subtypes (␣ 2A , ␣ 2B , and ␣ 2C ), each of which is activated by the endogenous catecholamines, epinephrine and norepinephrine, and performs multiple physiological functions via pertussis toxin-sensitive G i /G o proteins (1). Cellular signaling pathways regulated by ␣ 2A -AR in native cells include inhibition of adenylyl cyclase, activation of receptor-operated K ϩ channels, inhibition of voltage-gated Ca 2ϩ channels, and activation of the mitogen-activated protein kinase cascade (1)(2)(3). Many cells that express ␣ 2 -ARs are polarized, including renal and intestinal epithelia, where the ␣ 2 -AR serves to regulate sodium and water resorption (4,5), as well as neurons, where these receptors act to suppress neurotransmitter release (6). The physiological functions mediated by ␣ 2 -ARs in polarized cells are dependent upon precise localization of the receptor at the basolateral surface to gain access to neurally delivered and blood-delivered catecholamines.
The ␣ 2 -AR subtypes demonstrate unique targeting and retention profiles in polarized renal epithelial Madin-Darby canine kidney (MDCKII) cells in culture (7). Previous work in our laboratory has shown that the ␣ 2A -AR subtype is directly targeted to the basolateral surface, where it exhibits a half-life of 10 -12 h (8). Direct and exclusive basolateral targeting of ␣ 2A -AR was found to be dependent upon several noncontiguous regions within or near the bilayer, whereas retention of the receptor at the basolateral surface appears dependent upon the third intracellular (3i) loop (8). Deletion of the 3i loop results in accelerated surface turnover (t1 ⁄2 ϭ ϳ4.5 h) of the ␣ 2A -AR at the basolateral surface (9). Unlike the ␣ 2A -AR, the ␣ 2B -AR subtype is randomly targeted to both the apical and basolateral subdomains and then selectively retained at the basolateral surface of polarized MDCKII cells, where the receptor has a half-life comparable with that of the ␣ 2A -AR subtype (t1 ⁄2 ϭ ϳ10 -12 h) (7). Like for the ␣ 2A -AR, the 3i loop of the ␣ 2B -AR also is critical for basolateral surface stabilization of this subtype (10). In contrast to stable retention of the ␣ 2B -AR on the basolateral surface, the half-life on the apical surface is estimated to be dramatically shorter, on the order of minutes (7). Taken together, these data suggest that the stabilization/retention of ␣ 2A -and ␣ 2B -AR at specific membrane domains is most likely mediated through interactions of the 3i loop with other proteins either within or underlying the basolateral membrane surface and not present (or expressed at much lower density) at the apical surface.
The ␣ 2 -AR 3i loop has been used as a ligand to identify potential interacting proteins and has led to the identification of two ␣ 2 -AR-interacting molecules: 14-3-3 (11) and spinophilin (12). Spinophilin is an 817-amino acid, ubiquitously expressed, multidomain-containing protein with an apparent molecular mass on SDS-PAGE of ϳ130 kDa. It was originally identified both as a protein phosphatase 1 (PP1)-binding protein localized to dendritic spines, hence the name spinophilin (13), as well as an F-actin-binding protein (14). Spinophilin (also known as neurabin II) is highly homologous to the brainspecific protein, neurabin I (14). In addition to the domains described above, spinophilin contains a single PDZ (PSD-95, Discs large, ZO-1) domain and three coiled-coil domains at the C terminus, the latter of which mediate homo-multimerization in vitro (14) and may allow for the formation of multiprotein complexes in intact cells. Spinophilin previously was identified as a D2 dopamine receptor-interacting protein using the 3i loop of the D2 dopamine receptor as bait in a yeast two-hybrid screen (15). The D2 dopamine receptor-binding domain in spinophilin (residues 151-444), located between the F-actin-binding domain and the PP1 regulatory domain, also interacts with all three of the ␣ 2 -AR subtypes (12) and will be referred to as the receptor-interacting domain. Because reports in the literature, as well as our own observations, indicate that spinophilin is specifically enriched at the basolateral surface of polarized epithelial cells (12,15,16), we postulate that spinophilin may be involved in tethering and/or stabilizing the receptor at the cell surface via interactions with the ␣ 2 -AR 3i loops.
The present studies utilized two different biological systems to explore the role of spinophilin in ␣ 2 -AR stabilization at the cell surface. First, the unique targeting profile of the ␣ 2B -AR subtype in polarized MDCKII cells (random delivery with rapid turnover at the apical surface) was exploited to determine whether redirection of the receptor-interacting domain of spinophilin to the apical surface of polarized MDCKII cells would result in enhanced apical retention of randomly delivered ␣ 2B -AR. Second, the role of spinophilin in ␣ 2 -AR surface turnover was addressed by studying the internalization profile of the ␣ 2B -AR in mouse embryonic fibroblasts (MEFs) derived from wild type (Sp ϩ/ϩ ) or spinophilin knock-out (Sp Ϫ/Ϫ ) mice (17). The findings from both lines of investigation implicate spinophilin in the stabilization of the ␣ 2B -AR at the cell surface.

Materials
Transwell culture chambers (0.4-m pore size) were purchased from Costar (Cambridge, MA). Doxycycline hydrochloride was purchased from Sigma. NHS-SS-Biotin and Immunopure Immobilized Streptavidin were purchased from Pierce. Both [ 35 S]EasyTag™ Express Protein Labeling mix (1200 Ci/mmol) and [ 3 H]methoxy-inulin (126.5 mCi/g) were from PerkinElmer Life Sciences. Cysteine-and methionine-free DMEM was from Cellgro Mediatech. Dulbecco's modified Eagle's medium was prepared by the Cell Culture Core, a facility sponsored by the Diabetes Research and Training Center at Vanderbilt University Medical Center. Fetal calf serum was purchased from Sigma. Mouse monoclonal 12CA5 antibody against the HA epitope was obtained from BABCo. Affinity matrix-coupled high affinity rat monoclonal anti-HA antibody (clone 3F10), rat monoclonal anti-HA antibody (clone 3F10), and mouse monoclonal HA.11 antibody (clone 16B12) were purchased from Roche Applied Science. Mouse monoclonal anti-c-Myc (clone 9E10) ascites was purchased from Covance Research Products Inc. (Denver, PA). Both the mouse anti-gp135 and mouse anti-EGFR were gifts from Peter J. Dempsey (Department of Pathology, University of Washington, Harborview Medical Center, Seattle, WA). Rabbit anti-spinophilin antibody raised against spinophilin amino acids 286 -390 (18) was purified in our lab (for details see Ref. 12). Alexa Fluor 488-conjugated fluorescent goat anti-mouse, goat anti-rat, and goat anti-rabbit IgG were purchased from Molecular Probes (Eugene, OR). Cy3-conjugated donkey anti-mouse IgG was purchased from Jackson Immunochemicals. Sheep anti-mouse, donkey anti-rabbit, and goat anti-rat horseradish peroxidase-conjugated IgG were purchased from Amersham Biosciences. The rat p75 NTR cDNA was a generous gift from Dr. Bruce Carter (Department of Biochemistry, Vanderbilt University). The retroviral vector pBabe-HA-␣ 2B -AR was kindly provided by Drs. Dan Gil and John Donello (Allergan, Irvine, CA).

MDCKII Cell Culture and Polarization
MDCKII cells were plated at a density of 1.2 ϫ 10 6 cells/100-mm polycarbonate membrane filter (Transwell culture chambers, 0.4-m pore size) and cultured in DMEM supplemented with 10% fetal calf serum (Sigma) and 100 units/ml penicillin and 10 g/ml streptomycin at 37°C and 5% CO 2 with medium changes every other day for 5-7 days. Cells grown under these conditions achieve a morphologically and functionally polarized phenotype, as described previously (19). Leak assays of [ 3 H]methoxy-inulin were performed as described previously (20) prior to each half-life experiment to verify that the MDCKII cells had developed tight junctions and that the apical and basolateral compartments were functionally separated.

Generation of cDNAs Encoding Myc-p75-Spinophilin Fusion Proteins
The cDNA encoding the pTRE-Myc-p75-Sp151-483 fusion protein was generated via overlapping PCR extension using Pfu Turbo DNA polymerase (Stratagene). The Myc tag was inserted 5Ј to the coding start site of full-length rat p75 NTR and 3Ј of the N-terminal cleavable signal sequence. Four glycine residues were engineered via PCR onto the C terminus of p75 NTR with the intention of permitting independent folding of the spinophilin subdomain and decreased steric hindrance for interacting with other potential binding partners. The pTRE cDNA backbone (Clontech) has a tetracycline-inducible promoter that is intended to confer regulated expression of the fusion construct by treatment with the synthetic tetracycline analog, doxycycline. Two fusion proteins were generated. Myc-p75-Sp151-483 includes the receptorbinding domain and the PP1 regulatory domain of the full-length spinophilin, whereas Myc-p75-Sp151-586 also contains the PDZ-binding domain (cf. schematic of spinophilin domain structure in Fig. 2B). The cDNAs were sequenced in their entirety via 33 P-Thermo Sequenase Radiolabeled Terminator Cycle Sequencing Kit (U. S. Biochemical Corp.) to confirm that the sequences were correct.

Creation of Clonal MDCKII Cell Lines Expressing the Myc-p75-Spinophilin Fusion Protein
Permanent clonal cell lines were developed in MDCKII cells using FuGENE-6 (Roche Applied Science) transfection reagent, according to the manufacturer's protocol; 6 g of linearized pTET-On plasmid (Clontech) was co-transfected with 6 g of pTRE-Myc-p75-Spinophilin construct and 1 g of a vector encoding a hygromycin resistance gene into MDCKII cells already stably expressing the HA-␣ 2B -AR, generated as described previously (7). Stably transfected cells were selected through growth in 400 g/ml hygromycin and assayed for HA-␣ 2B -AR expression via radioligand binding analysis using the antagonist [ 3 H]rauwolscine (21) and for Myc-p75-spinophilin expression and inducibility via Western analysis and immunofluorescence using anti-c-Myc antibody. Despite the use of an inducible expression system, Myc-p75-spinophilin expression occurred even in the absence of doxycycline. Nonetheless, the cells were treated overnight with 1 g/ml doxycycline before the day of the experiment to assure maximal Myc-p75-spinophilin expression.

Immunofluorescence in Polarized MDCKII Cells
Polarized MDCKII cells stably expressing HA-␣ 2B -AR were grown on 12-mm Transwells for 5-7 days (as described above) and processed as described previously for immunolocalization of HA-␣ 2B -AR and endogenous spinophilin (12) and for detection of endogenous apical (gp135) and basolateral (EGFR) marker proteins (22) (see Fig. 1). MDCKII cells expressing Myc-p75-Sp151-483 fusion protein in the HA-␣ 2B -AR background were treated with 1 g/ml doxycycline overnight to maximize the expression of Myc-p75-Sp151-483 prior to staining. All of the steps were performed essentially as described previously (22) except that a rat anti-HA (clone 3F10) antibody (Roche Applied Science) diluted 1:1000 in blocking buffer was used for the detection of HA-␣ 2B -AR (see Fig. 2D).

Immunoprecipitation of Apically Targeted Myc-p75-Spinophilin Fusion Proteins from Stable MDCKII Cell Lines
MDCKII cells lines stably expressing the HA-␣ 2B -AR alone or Myc-p75-Sp151-483 or Myc-p75-Sp151-586 in the HA-␣ 2B -AR background were grown to confluence in 100-mm tissue culture dishes. All of the dishes were treated with 1 g/ml doxycycline for 16 h before harvesting the cells to maximize fusion protein expression. On the day of the assay, the dishes were washed twice with Dulbecco's phosphate-buffered saline supplemented with 1 mM MgCl 2 and 0.5 mM CaCl 2 (DPBS/CM) (4°C) and scraped into 12 ml of lysis buffer (15 mM Tris-HCl, 5 mM EGTA, 5 mM EDTA, pH 7.6, with N-methyl-D-glucosamine) with protease inhibitors (1 g/ml soybean trypsin inhibitor, 0.5 g/ml leupeptin, 100 M phenylmethylsulfonyl fluoride). The cells were passaged through a 20-gauge needle 10 times and then centrifuged for 20 min at 39,000 ϫ g. The supernatant was aspirated, and the pellet was resuspended in 500 l of radioimmune precipitation buffer (RIPA) (150 mM NaCl, 50 mM Tris-HCl, pH 8.0, 5 mM EDTA, 1% Nonidet P-40, 0.5% deoxycholate, 0.1% SDS) plus protease inhibitors (same as above) with five passages through a 20-gauge needle followed by 10 passages through a 25-gauge needle. The detergent-extracted membranes were cleared of RIPA-insoluble debris by centrifugation for 1 h at 100,000 ϫ g. The supernatants were precleared for 30 min with rotation at 4°C with protein G-agarose beads pre-equilibrated in RIPA buffer. Anti-c-Myc antibody (1:100 dilution) was added to each of the precleared samples and incubated overnight at 4°C with rotation. The next day, 30 l of a 1:1 slurry of protein G-agarose (pre-equilibrated in RIPA buffer containing 2.5 mg/ml bovine serum albumin (BSA)) was added to each tube and incubated for 2 h at 4°C with rotation. The protein G-agarose was pelleted and washed four times with 1 ml of ice-cold RIPA buffer plus protease inhibitors. Immunoisolated protein from the protein Gagarose was eluted by two sequential incubations for 10 min each with 25 l of 1ϫ SDS sample buffer (50 mM Tris-HCl, pH 8.0, 2% SDS, 10% glycerol, 100 mM dithiothreitol, 0.1% bromphenol blue) at 70°C. The eluates were pooled, and the entire sample was resolved by 10% SDS-PAGE and transferred to nitrocellulose for Western blot analysis as described previously (23). Myc-p75-Sp151-483 or Myc-p75-Sp151-586 was detected by incubation with rabbit anti-Sp286 -390 antibody (1: 1000) followed by donkey anti-rabbit horseradish peroxidase-conjugated secondary antibody (1:2000) and visualized by ECL (Amersham Biosciences).

Determination of Protein Half-life on the Apical Surface of Polarized MDCKII Cells
HA-␣ 2B -Adrenergic Receptors-A metabolic radiolabeling strategy was used to determine receptor half-life of ␣ 2B -AR at the apical surface of polarized MDCKII cells because of the low steady state density of the ␣ 2B -AR on this surface (7). Cells grown 5-7 days in 100-mm Transwell culture were treated overnight with 1 g/ml doxycycline. The day of the assay the cells were washed once in DPBS/CM and then, simultaneous with performing a transepithelial [ 3 H]methoxy-inulin leak assay (see above), incubated for 2 h at 37°C in serum-free, cysteine/methioninefree DMEM. The cells were then pulsed for 45 min with 900 l of 2 Ci/l [ 35 S]cysteine/methionine in cysteine/methionine-free DMEM at 37°C and 5% CO 2 . "Chasing" of the metabolically labeled cells was achieved by adding serum-free DMEM containing 1 mM cysteine and 1 mM methionine (chase medium) (9 ml apical/9 ml basolateral) and returning the dishes to 37°C for 40 min to allow delivery of all of the receptor labeled during the pulse phase to the cell surface.
To begin the determination of the surface half-life of metabolically labeled receptors, the apical surface of polarized MDCKII cells was biotinylated as follows. Transwells were washed once with 4°C DPBS/CM and transferred to the cold room on ice. After washing, the Transwells were equilibrated for 10 min in 4°C TEA buffer (250 mM sucrose, 2 mM CaCl 2 , 2 mM MgCl 2 , 10 mM triethanolamine, pH 9.0) and then biotinylated for 20 min at 4°C on the apical surface by incubating freshly made 1 mg/ml sulfo-NHS-SS-Biotin in TEA buffer in the apical chamber (an equal volume of TEA buffer without sulfo-NHS-SS-Biotin was also added to the basolateral chamber). The biotinylation step was repeated to assure quantitative labeling of the receptor. Washing the cells with 100 mM glycine in DPBS/CM for 10 min quenched the biotinylation reaction. After washing twice more with 4°C DPBS/CM and once with 4°C serum-free DMEM chase medium, the cells were transferred to 37°C chase medium and returned to the 37°C incubator for the indicated times.
At varying time points, the amount of biotinylated HA-␣ 2B -AR on the apical surface was quantified by sequential immunoisolation and streptavidin chromatography. Selected dishes were washed twice for 10 min at 4°C with DPBS/CM, scraped into 12 ml of lysis buffer (15 mM Tris-HCl, 5 mM EGTA, 5 mM EDTA, pH 7.6, with N-methyl-D-glucosamine) containing protease inhibitors (1 g/ml soybean trypsin inhibitor, 0.5 g/ml leupeptin, 100 M phenylmethylsulfonyl fluoride), triturated 10 times through a 20-gauge needle, and then centrifuged at 39,000 ϫ g for 20 min. The pellet was resuspended in 1 ml of RIPA buffer plus protease inhibitors and incubated on ice 30 min before centrifugation for 1 h at 100,000 ϫ g. Supernatants from this centrifugation (the solubilized preparation) were incubated with 25 l of a 1:1 slurry of pre-equilibrated rat anti-HA affinity matrix overnight at 4°C with rotation. The affinity matrix was pelleted and washed four times with 1 ml of ice-cold RIPA buffer plus protease inhibitors before elution of the immunoisolated HA-␣ 2B -AR from the affinity matrix by incubation twice for 10 min with 100 l of SDS sample buffer (1.6% SDS, 8.3% glycerol, 167 mM Tris, pH 8.0) at 70°C. The eluates were pooled and brought to 1.5 ml with RIPA (containing no SDS) plus protease inhibitors. The sample was allowed to sit for 10 min at room temperature to equilibrate the component detergents.
To isolate the apically biotinylated HA-␣ 2B -AR from the entire immunoisolate, streptavidin chromatography was performed as follows. The samples were incubated with a 1:1 slurry of streptavidin-agarose (50 l) pre-equilibrated in RIPA buffer for 2 h at 4°C with rotation. Pelleted streptavidin-agarose was washed three times with 1 ml of ice-cold RIPA buffer containing protease inhibitors, and biotinylated HA-␣ 2B -AR was eluted by incubation twice for 20 min with 100 l of SDS sample buffer containing 50 mM dithiothreitol at 90°C. The eluted samples were then incubated 40 min at 50°C. N-Ethylmaleimide was added to a final concentration of 15 mM, and the samples were incubated for an additional 40 min at 50°C. The rationale for the high dithiothreitol/N-ethylmaleimide treatment is to alkylate all sulfhydryl residues, thus better resolving the ␣ 2 -AR preparation on SDS-PAGE (24).
The samples were resolved overnight for a total of 160mAmp-hr on a 7.5-20% gradient SDS-polyacrylamide gel. The gels were treated with En[ 3 H]ance Intensifying Solution (PerkinElmer Life Sciences) according to the manufacturer's protocols, dried, and exposed to BioMAX MR film. The bands on the film were quantitated using SCION image software, and/or bands were cut from the gel and counted directly in scintillation mixture. Equivalent findings were obtained from either quantitation procedure.
Endogenous gp135-Because endogenous gp135 is expressed at a relatively high concentration on the apical surface of polarized MDCKII cells, the surface half-life of this protein was determined by surface biotinylation, extraction into RIPA at various time points, resolution by SDS-PAGE, and identification of biotinylated gp135 via Western blot analysis for gp135, using methods described previously (23).

Culturing of MEFs
A 13.5-day pregnant female mouse (Sp ϩ/ϩ or Sp Ϫ/Ϫ ) (17) was sacrificed, and the embryos were collected. The soft, dark colored tissues (i.e. heart, liver, and spleen) were dissected away from the embryo, and the head was removed. The remaining tissue was transferred to the barrel of a 5-ml syringe (five embryos/syringe) and passed through an 18gauge needle into 3 ml of DPBS. The tissue was further dissociated by trituration five times, and the cell suspension was transferred to a 150-mm culture dish containing 25 ml of complete medium (DMEM with 10% fetal calf serum, 100 units/ml penicillin, and 10 g/ml streptomycin supplemented with 2 mM glutamine). The cells were grown at 37°C and 5% CO 2 until the plates reached confluency, at which point the cells were split 1:5, expanded to confluency, and frozen at 2 ϫ 10 6 cells/ml in freezing medium (50% fetal calf serum, 12% Me 2 SO in DMEM).

Transduction of MEFs with a Retroviral Vector
Encoding HA-␣ 2B -AR Primary cultures of MEFs (Sp ϩ/ϩ or Sp Ϫ/Ϫ ) (17) were seeded at 1.4 ϫ 10 6 cells/100-mm dish the day before transduction with 4 ml of one part retroviral supernatant containing HA-␣ 2B -AR-encoding virions harvested from BOSC cells (25) and one part complete DMEM containing a final concentration of 12 g/ml polybrene. The viral application was repeated four times over the course of 8 h at 37°C and 5% CO 2 , empirically determined to yield optimal transduction. Afterward, the cells were returned to 9 ml of complete medium. Three days posttransduction, the cells were assayed for HA-␣ 2B -AR expression via radioligand binding analysis, essentially as described previously (9). For specifically indicated experiments, the transduced cells were selected for retroviral vector expression by treatment overnight with 4 g/ml puromycin (the pBabe retroviral vector carries the resistance gene for puromycin).

Measuring Turnover of the HA-␣ 2B -AR in MEFs
Intact Cell ELISA Assay-The day before the assay, Sp ϩ/ϩ or Sp Ϫ/Ϫ MEFs (selected for HA-␣ 2B -AR expression in 4 g/ml puromycin) were plated on poly-D-lysine-coated 96-well culture plates at a density of 4 ϫ Reversible Biotinylation-Surface HA-␣ 2B -AR was labeled on ice with a disulfide cleavable biotin (sulfo-NHS-SS-Biotin), stimulated by agonist at 37°C, and then treated with the cell-impermeant reducing agent, 2-mercaptoethanesulfonic acid (MESNA). Receptors that have trafficked to the inside of the cell at the time of MESNA treatment are protected from reduction and can be subsequently isolated via streptavidin agarose. Total receptor available at the cell surface at time 0 (t 0 ) was defined by the difference in the amount of biotinylated receptor detected in the absence of MESNA treatment and that detected following immediate reversal by MESNA (thereby revealing the quantity of surface receptor biotinylation that MESNA can reverse efficiently). Internalized receptor detected at various time points (represented by the MESNA-insensitive fraction) was quantitated following Western analysis and expressed as a fraction of the total surface receptor available at t 0 .
For each experiment, Sp ϩ/ϩ or Sp Ϫ/Ϫ MEFs expressing HA-␣ 2B -AR were plated at 1 ϫ 10 6 cells/60-mm dish the day before the assay (1 dish for each time point assayed) and treated overnight with phentolamine (1 M) as described above. The day of the assay, the cells were washed three times for 10 min at 37°C with serum-free DMEM containing 0.1% BSA, placed on ice for 20 min, and then washed once with 4°C DPBS/ CM. The cells were incubated at 4°C for 30 min with 100 g/ml sulfo-NHS-SS-Biotin to label surface proteins. Following biotinylation, the cells were washed twice with DPBS/CM at 4°C and once with 4°C serum-free DMEM. Next, the cells were stimulated for the indicated times by incubation at 37°C in serum-free DMEM containing epinephrine (100 M) and propranolol (1 M). The incubation was terminated by replacement of the 37°C medium with 4°C DPBS/CM, followed by two 20-min incubations at 4°C with MESNA (250 mM) in DPBS/CM. Following incubation with MESNA, the cells were washed once with 4°C DPBS/CM, and residual MESNA was quenched by incubation with 5 mg/ml iodoacetamide for 20 min at 4°C. The cells were then washed once more in DPBS/CM, scraped into 100 l of D␤M-CHS extraction buffer (4 mg/ml dodecyl ␤-maltoside, 0.8 mg/ml cholesteryl hemisuccinate, 20% glycerol, 25 mM glycylglycine, 20 mM HEPES, 100 mM NaCl, 5 mM EGTA, 0.1 mM phenylmethylsulfonyl fluoride, 10 units/ml aprotinin) and triturated five times with a 25-gauge needle. Supernatant of a 1-h microcentrifugation at 12,000 rpm at 4°C was incubated with a 1:1 slurry (40 l) of streptavidin-agarose (pre-equilibrated in D␤M-CHS extraction buffer containing 2.5 mg/ml BSA) for 1 h with rotation at room temperature. The streptavidin-agarose pass-through was saved, and the strepavidin-agarose resin was washed three times with 500 l of ice-cold D␤M-CHS wash buffer (0.5 mg/ml dodecyl ␤-maltoside, 0.1 mg/ml cholesteryl hemisuccinate, 25 mM glycylglycine, 20 mM HEPES, 100 mM NaCl, 5 mM EDTA). The biotinylated HA-␣ 2B-AR was eluted twice at 90°C into 30 l of SDS sample buffer (50 mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerol, 0.1% bromphenol blue) containing 50 mM dithiothreitol, resolved by 10% SDS-PAGE, and transferred to nitrocellulose for Western blot analysis as described previously (23).

Co-localization of Endogenous Spinophilin with HA-␣ 2B -AR
in Polarized MDCKII Cells-Immunofluorescence staining and confocal laser scanning microscopy were used to confirm the localization of both the HA-␣ 2B -AR and endogenous spinophilin in the MDCKII cell line used for these studies (Fig. 1). As has been shown previously, the HA-␣ 2B -AR is localized to the basolateral surface in these polarized epithelial cells, albeit with a small pool of presumably endocytosed ␣ 2B -AR (7). Endogenous spinophilin also is enriched at the basolateral surface, as has been suggested in the literature (12,14,15). The co-localization of these two proteins at the basolateral surface is indicated by the yellow signal present in the overlay of the two images, particularly evident in the z scan. We postulated that the basolateral localization of spinophilin in MDCKII cells may contribute to retention of ␣ 2 -AR subtypes at that surface and that the lack of apical spinophilin may explain the rapid apical turnover of randomly delivered ␣ 2B -AR.
Redirection of a Spinophilin Subdomain to the Apical Surface of MDCKII Cells-To test the hypothesis that spinophilin contributes to cell surface ␣ 2 -AR retention, we explored whether redirection of the receptor-interacting domain of spinophilin to the apical surface of polarized MDCKII cells would lead to enhanced apical retention of randomly delivered ␣ 2B -AR. Fig. 2B provides a schematic diagram of the domain structure of the spinophilin protein (14). A region of spinophilin sequence (amino acids 151-483) including the receptor-interacting domain was fused-in frame to Myc epitope-tagged p75 NTR , a single transmembrane spanning protein known to be expressed predominantly (80%) at the apical surface of polarized MDCKII cells (26). The spinophilin sequences were separated from p75 NTR via a tetraglycine linker to permit independent folding and accessibility of the spinophilin domains. A cDNA encoding this fusion protein was then stably expressed in MDCKII cells already stably expressing HA-␣ 2B -AR, as described under "Experimental Procedures." Expression of the Myc-p75-Sp151-483 and Myc-p75-Sp151-586 fusion proteins in these HA-␣ 2B -AR-expressing MDCKII cell lines was confirmed via immunoprecipitation and Western blotting. As can be seen in Fig. 2C, the Myc-tagged fusion proteins can be specifically enriched by immunoprecipitation of FIG. 1. Localization of ␣ 2B -AR and endogenous spinophilin in polarized MDCKII cells. Immunolocalization of ␣ 2B -AR in polarized MDCKII cells was achieved using an antibody directed against the HA epitope (see "Experimental Procedures") and reveals a basolateral staining pattern at steady state. The localization of endogenous spinophilin reveals a similar basolateral staining pattern. Co-localization of the two proteins is indicated by the presence of a yellow signal resulting from overlay of the two images (third panel).
detergent-solubilized membrane fractions with anti-c-Myc antibody and then identified on Western blots with an antibody against an epitope within spinophilin (amino acids 286 -390) (lanes 2 and 3). As expected, no Myc-spinophilin fusion protein is detected in the parental cell line expressing only heterologous HA-a 2B -AR (lane 1).
As shown in Fig. 2D, immunolocalization studies using an antibody against the Myc epitope reveal that Myc-p75-Sp151-483 is expressed at the apical surface of polarized MDCKII cells (Fig. 2D, first panel) in a manner similar to the expression pattern of the known endogenous apical marker protein, gp135 (Fig. 2D, third panel). A similar apical expression pattern also FIG. 2. Redirection of the receptor-interacting domain of spinophilin to the apical surface of polarized MDCKII cells. A, schematic diagram reflecting the targeting and retention profile of the ␣ 2B -AR subtype in polarized MDCK II cells. The ␣ 2B -AR is delivered randomly to both the apical and basolateral surfaces. The half-life of ␣ 2B -AR at the apical surface is dramatically shorter (t1 ⁄2 Յ 60 min) than its half-life at the basolateral surface (ϳ10 -12 h), which explains the basolateral enrichment of ␣ 2B -AR observed in steady state assessments (cf. Fig. 1). B, schematic diagram of wild type spinophilin and the spinophilin fusion protein used to target spinophilin subdomains to the apical surface of polarized MDCKII cells. The apically targeted spinophilin fusion protein was created by cloning the DNA encoding spinophilin amino acids 151-483 in frame to the Myc epitope-tagged, apically targeted p75 NTR . This structure contains the receptor-interacting domain as well as the PP1 regulatory domain of spinophilin. A tetraglycine linker was engineered between the p75 NTR and the spinophilin subdomain to ensure independent folding of the separate domains. C, immunoisolation of apically targeted Myc-p75-spinophilin fusion proteins from stable MDCKII cell lines. The membranes were harvested from MDCKII cells stably expressing HA-␣ 2B-AR alone (lane 1) or HA-␣ 2B -AR along with Myc-p75-Sp151-483 (lane 2) or Myc-p75-Sp151-586 (lane 3), solubilized in RIPA buffer, and subjected to immunoprecipitation (IP) with anti-c-Myc monoclonal antibody (clone 9E10). Samples eluted from protein G-agarose beads were separated via 10% SDS-PAGE, transferred to nitrocellulose, and blotted with the polyclonal rabbit anti-Sp286 -390 antibody. The small difference in molecular mass of these large fusion proteins was not readily detectable following PAGE in 10% gels. The bands were visualized using ECL (see "Experimental Procedures" for details). D, apical localization of the Myc-tagged p75-spinophilin subdomain is verified by immunofluorescent localization and laser scanning confocal microscopy. Fixed cells were permeabilized and stained for the indicated endogenous marker proteins, EGFR (basolateral) and gp135 (apical), (third and fourth panels, respectively), or for the exogenously expressed Myc-p75-Sp151-483 fusion protein either alone (first panel), or overlaid with HA-␣ 2B -AR (second panel) (see "Experimental Procedures" for details). The xy plane shows the apical surface of the cells, whereas the z plane shows a lateral slice through the cells along the line depicted in yellow in the xy plane. Immunolocalization of the Myc-p75-Sp151-483 fusion protein was achieved using an antibody against the c-Myc epitope. was detected for the clonal cell line expressing Myc-p75-Sp151-586 (data not shown). Fig. 2D (second panel) shows that the apical expression pattern of Myc-p75-Sp151-483 is in marked contrast to the basolateral expression pattern of endogenous spinophilin, heterologous HA-␣ 2B -AR at steady state (Fig. 1), or the EGFR (Fig 2D, fourth panel), a marker protein for the basolateral surface. Thus, our fusion protein strategy successfully delivered the receptor-interacting domain of spinophilin to the apical surface and did so without altering expression of endogenous markers for the polarized phenotype, gp135, and the EGFR.
An Apically Targeted Spinophilin Subdomain Extends the Apical Half-life of Randomly Delivered ␣ 2B -AR-To determine whether apical expression of the receptor-interacting domain of spinophilin would extend the apical surface half-life of randomly delivered ␣ 2B -AR, a cell surface biotinylation strategy in metabolically labeled, polarized MDCKII cells was used to quantify the loss of apical HA-␣ 2B -AR over time (see Refs. 7 and 19 and "Experimental Procedures"). The autoradiogram in Fig.  3A shows that the rapid loss of apical HA-␣ 2B -AR over the initial 60 min is attenuated in cells expressing Myc-p75-Sp151-483. Fig. 3B summarizes data from multiple experiments and shows a time course extended over 6 h. These data reveal a biphasic loss of HA-␣ 2B -AR from the apical surface for cells not expressing exogenous apical spinophilin, such that ϳ50% of the apical HA-␣ 2B -AR was lost with a t1 ⁄2 ϭ ϳ50 min. By contrast, in cells expressing the apically targeted spinophilin fusion protein, the half-life of the HA-␣ 2B -AR at the apical surface is extended to ϳ3.6 h. Coincidentally, in cells expressing Myc-p75-Sp151-483, HA-␣ 2B -AR loss from the surface occurred at a constant rate, similar to that of the slower phase observed in cells not expressing exogenous apical spinophilin. The initial phase on the apical surface was comparably lengthened in a cell line expressing a p75 fusion protein that also contains the PDZ domain of spinophilin, Myc-p75-Sp151-586 (data not shown), indicating that protein interactions fostered by the PDZ domain do not contribute to ␣ 2B -AR retention, at least on the apical surface. Moreover, because these spinophilin fusion constructs did not contain coiled-coil or F-actin-binding domains, neither multimerization nor F-actin-binding of spinophilin appears to be necessary to stabilize ␣ 2B -AR expression at the cell surface. In addition, the extended apical half-life of the HA-␣ 2B -AR does not appear to be a consequence of generalized changes in apical membrane turnover, because the turnover of the endogenous apical marker protein, gp135, was not significantly altered in cells expressing Myc-p75-Sp151-483 (Fig. 3C).
Previous studies of apical retention of the randomly delivered HA-␣ 2B -AR had only revealed a single population of rapidly (t1 ⁄2 ϭ ϳ15-45 min) turning over HA-␣ 2B -AR (7). The second, slower phase of the HA-␣ 2B -AR loss from the apical surface detected in the present studies is not due to leak of the biotinylating reagent to the basolateral compartment, because the slower phase of HA-␣ 2B -AR loss from the apical surface (t1 ⁄2 ϭ ϳ3.6 h) is entirely different from the half-life of the HA-␣ 2B -AR on the basolateral surface (t1 ⁄2 ϭ ϳ10 -12 h). Our lack of detection of the second, extended apical expression (t1 ⁄2 ϭ ϳ3.6 h) of a subfraction of the apically delivered ␣ 2B -AR in these earlier studies may have been due to the lesser amount of biological material evaluated (the present studies examined cells grown in 100-mm rather than 24-mm Transwell cultures). It is not entirely surprising that the apical half-life of the HA-␣ 2B -AR in cells engineered to express an apically targeted receptor-interacting domain of spinophilin is extended beyond 60 min to ϳ3.6 h but is not restored to that of the half-life of the HA-␣ 2B -AR at the basolateral surface (t1 ⁄2 ϭ ϳ10 -12 h); it is likely that interactions of spinophilin and/or the receptor with additional proteins underlying the basolateral surface also contribute to ␣ 2B -AR residence time. For example, although the F-actinbinding and coiled-coil domains are not necessary for stabilization of ␣ 2B -AR expression on the apical surface (see above), multimerization of endogenous spinophilin and/or its inter- action with F-actin may contribute to the more extended half-life on the basolateral surface. Despite the enhanced apical retention of HA-␣ 2B -AR in cells harboring the apically targeted spinophilin fusion protein observed in biochemical studies, we were unable to detect an apical pool of HA-␣ 2B -AR via immunofluorescence studies, perhaps because metabolic labeling and biotinylation are intrinsically more sensitive for examining receptor turnover at the cell surface than is immunofluorescence.
The most straightforward interpretation of our findings is that the receptor-interacting domain of spinophilin itself is stabilizing the ␣ 2B -AR at the apical surface. However, we cannot rule out a role for the PP1 interacting domain, which is inherent in each of the fusion proteins examined.
Agonist-induced Internalization Is Enhanced in Spinophilindeficient (Sp Ϫ/Ϫ ) MEFs-As a complementary strategy to explore whether spinophilin plays a role in stabilizing the ␣ 2B -AR at the cell surface, we took advantage of primary MEFs derived from wild type (Sp ϩ/ϩ ) or spinophilin knock-out (Sp Ϫ/Ϫ ) mice (17). Western blot analysis of whole cell lysate using an antibody against endogenous spinophilin reveals that spinophilin is absent from the Sp Ϫ/Ϫ MEFs, whereas it is readily detected in the wild type (Sp ϩ/ϩ ) cells (Fig. 4A).
Wild type (Sp ϩ/ϩ ) and spinophilin knock-out (Sp Ϫ/Ϫ ) MEFs were transduced with HA-␣ 2B -AR using a retroviral expression system, and receptor expression was verified via radioligand binding analysis. In contrast to the lack of effect of agonists on short term ␣ 2A -AR turnover in MDCKII (8) or other (27,28) target cells, the ␣ 2B -AR subtype has previously been documented to rapidly internalize in response to agonist treatment (28 -31). We postulated that if spinophilin is important for stabilizing the receptor at the cell surface, then internalization of the ␣ 2B -AR may be accelerated in a cell background lacking spinophilin. We evaluated HA-␣ 2B -AR internalization using two independent strategies in Sp ϩ/ϩ versus Sp Ϫ/Ϫ MEFs: 1) cell surface ELISA, which measured the loss of receptor from the cell surface, and 2) reversible biotinylation, which examined FIG. 4. Mouse embryo fibroblasts lacking spinophilin (Sp ؊/؊ ) reveal an accelerated turnover of HA-␣ 2B -AR when compared with wild type (Sp ؉/؉ ) MEFs. A, endogenous spinophilin is readily detected in whole cell lysates of Sp ϩ/ϩ , but not Sp Ϫ/Ϫ , MEFs. A single 60-mm dish containing Sp ϩ/ϩ or Sp Ϫ/Ϫ MEFs was harvested into 500 l of RIPA buffer plus protease inhibitors (cf. "Experimental Procedures") to create a solubilized extract. Increasing volumes of this extract were resolved on a 10% SDS mini-gel. The Western blot (WB) was probed for endogenous spinophilin using rabbit anti-Sp286 -390 antibody. B, HA-␣ 2B -AR loss from the cell surface was quantitated using an intact cell ELISA. Sp ϩ/ϩ or Sp Ϫ/Ϫ MEFs expressing HA-␣ 2B -AR were stimulated at 37°C with ␣ 2 -AR agonist, fixed, and then incubated with antibody against the HA epitope. The amount of antibody detected on the cell surface at the indicated times was quantitated via a colorimetric substrate for a secondary antibody-conjugated enzyme (see "Experimental Procedures"). The values shown are the means Ϯ S.E. for n ϭ 3 or more individual experiments performed with four to six replicates/experiment. After 60 min of agonist stimulation, ϳ55% of the HA-␣ 2B -AR had been lost from the cell surface in the Sp Ϫ/Ϫ MEFs, whereas in the Sp ϩ/ϩ MEFs only ϳ30% of the HA-␣ 2B -AR was lost. The values were determined using an unpaired t test. *, p ϭ 0.005; **, p ϭ 0.0001. In the absence of agonist, little or no HA-␣ 2B -AR is lost from the cell surface (data not shown).

FIG. 5.
Reversible biotinylation reveals the more rapid loss of HA-␣ 2B -AR from the cell surface of Sp ؊/؊ MEFs than from Sp ؉/؉ MEFs. Receptor loss from the cell surface was quantitated using a reversible biotinylation protocol as described under "Experimental Procedures." Internalized receptor (with MESNA) at each time point was determined as a percentage of total biotinylated receptor available at time 0. A, raw data from a representative experiment showing that the amount of MESNA-insensitive (internalized) HA-␣ 2B -AR increases over the course of agonist stimulation to a greater extent in the Sp Ϫ/Ϫ MEFs than in the Sp ϩ/ϩ MEFs. B, band densities resulting from the autoradiography exposures from multiple experiments (Sp ϩ/ϩ , n ϭ 9; Sp Ϫ/Ϫ , n ϭ 8) were quantified by scanning the images from the film and measuring the relative pixel intensities using SCION Image software. The value determined for t 0 ϩ MESNA was subtracted as background from the values obtained from each of the time points assayed and then expressed as a percentage of the value determined for t 0 Ϫ MESNA. The computer-generated curve represents a statistically significant difference between the internalization of HA-␣ 2B -AR in Sp ϩ/ϩ and Sp Ϫ/Ϫ MEFs. *, p ϭ 0.0461. HA-␣ 2B -AR that is internalized and protected from MESNAevoked removal of the biotin moiety from the cell surface.
For the intact cell ELISA, Sp ϩ/ϩ versus Sp Ϫ/Ϫ MEFs were treated with an ␣ 2 -AR agonist for the indicated times and then labeled with primary antibody directed against the HA epitope. Consistent with previous findings in cultured cell lines (all of which express endogenous spinophilin) (28,31), ϳ30% of the HA-␣ 2B -AR was lost from the cell surface of Sp ϩ/ϩ MEFs in response to agonist treatment for 60 min (Fig. 4B). In the absence of agonist, the HA-␣ 2B -AR remained at the cell surface for the entire time course of the incubation (data not shown). Importantly, however, ϳ55% of the HA-␣ 2B -AR was lost from the cell surface of Sp Ϫ/Ϫ MEFs over the same 60-min incubation duration (Fig. 4B). These data provide additional evidence that the presence of spinophilin stabilizes the ␣ 2B -AR at the cell surface.
A reversible biotinylation strategy, which measures the amount of internalized HA-␣ 2B -AR that occurred over time following agonist exposure, was also exploited. For these studies, HA-␣ 2B -AR expressing MEFs (Sp ϩ/ϩ and Sp Ϫ/Ϫ ) were labeled at 4°C with the membrane-impermeant, cleavable biotinylating reagent, sulfo-NHS-SS-biotin, before incubation with agonist for varying amounts of time. At the end of each incubation period, the cells were placed at 4°C and treated with a membrane-impermeant reducing agent, MESNA, to cleave disulfide-linked biotin remaining on the surface. Receptors isolated via streptavidin-agarose from detergent-solubilized cells represent the receptors that remain biotinylated at each incubation point after MESNA treatment, which reveals the fraction of the receptor pool that was protected from reversal of biotinylation (i.e. internalized) during the course of the experiment. As shown in Fig. 5, this experimental strategy also reveals increased internalization of the HA-␣ 2B -AR following agonist treatment in Sp Ϫ/Ϫ MEFs as compared with Sp ϩ/ϩ MEFs.
Conclusion-The present studies demonstrate a role for spinophilin in the stabilization/retention of the ␣ 2B -AR at the cell surface both in a cultured cell system and in cells derived from spinophilin knock-out mice. An apically targeted spinophilin subdomain containing the receptor-interacting domain extends the half-life of randomly delivered HA-␣ 2B -AR at the apical surface of polarized MDCKII cells, where the receptor exhibits a rapid surface turnover under wild type conditions in contrast to its long-lived retention (t1 ⁄2 ϭ ϳ10 -12 h) at the basolateral surface. Presumably, the transient stabilization of the HA-␣ 2B -AR at the apical surface caused by redistribution of the receptor-interacting domain of spinophilin to that surface reflects the role that endogenous spinophilin plays in basolateral stabilization of all three ␣ 2 -AR subtypes, because all three interact with spinophilin via their 3i loops (12). This interpretation is further supported by studies utilizing mouse embryonic fibroblasts derived from spinophilin knock-out mice in which two distinct lines of evidence were each consistent with enhanced agonist-induced internalization of HA-␣ 2B -AR in Sp Ϫ/Ϫ cells as compared with wild type Sp ϩ/ϩ cells expressing endogenous spinophilin. Taken together, these data suggest that spinophilin plays a role in stabilizing the receptor at the cell surface. The precise mechanism for this stabilization is not known but could involve physically anchoring the receptor to the actin cytoskeleton, as is seen with the D2 dopamine receptor and actin-binding protein-280 (ABP-280, also known as Filamin-A) (32,33). Alternatively, spinophilin may impede in-teraction with other molecules, such as ␤-arrestin (34), that would foster receptor internalization. Our interpretation that spinophilin stabilizes the ␣ 2 -AR at the surface likely extends to other G i /G o -coupled G protein-coupled receptors, because spinophilin also interacts with the 3i loop of the D2 dopamine receptor (15). Future studies can establish whether the multidomain nature of spinophilin may serve to bring other proteins into the receptor microcompartment that participate not only in receptor localization but also in coordination of receptor-elicited signal transduction.